About this video
The video explains why adding a string and a number causes a type error and how to fix it with type casting. It shows the str, int and float functions, explains that input always returns a string, shows why int fails on a value like "89.5" (a value error) so float is the better choice, and notes that casting a float to an integer chops off the decimal part.
You'll be able to
- explain why mixing strings and numbers in + causes a type error.
- use str, int and float to cast values to other data types.
- explain why the output of input usually needs casting before doing math.
- predict when casting causes a value error and what int does to a float.
Key words
- Type error
- An error caused by using values of data types that don't work together, like a string plus a float.
- Type casting
- Converting a value from one data type to another.
- str function
- Casts a value to a string, as if putting quotation marks around it.
- int function
- Casts a value to an integer.
- float function
- Casts a value to a float.
- Value error
- An error when a value can't be cast in any meaningful way, like int("banana").
Transcript
- [Instructor] Have you ever tried to make your print output a bit more descriptive like this, only to get a type error? Why does that happen and how do we fix it? Let's put our debugging skills to work. We saw that my program last worked when I was just printing the average score.
It's only when I tried to add this string at the front that I started getting an error. This error message tells me the computer's having trouble concatenating a string with a float. My average quiz score is a string, and the variable average score contains the value 89.5, which is a float, and we know that computers can't add strings and floats together.
They can only concatenate strings and strings or add some combination of integers and floats. Okay, so we know why we're getting that error, but what can we do about it? We, as the programmers, need to help the computer out here. The computer is very stubborn about its data types matching, and we can't change that, but we can change our values to be different data types.
This is called type casting where we convert or cast a value from one data type to another data type. To cast a value to a string, we use the string function. Remember that a function takes in an input value, performs some actions on it, and then returns out an output value.
With the string function, our input value is the value that we want to cast to a string, and our output value is that input value as a string. Casting a float or an integer to a string is pretty straightforward. It's just as if we put quotation marks around it.
Back to our program, all we have to do is call the string function on average score. Then that float value 89.5 that the average score variable contains, will be cast to the string 8 9.5. If we ignore the variable for a second, this is the same as saying plus 89.5, which does give a type error versus string 8 9.5, which doesn't give a type error.
Once we have this expression with two strings, the computer can just concatenate them. What if I wanna see how my next quiz score will affect my average? I might add an input prompt here to let me enter in different scores. So I update my average score calculation to include that new score, and now I'm getting a type error again.
That's because the output value that the input function returns is always a string even if what the user typed kind of looks like a number. That means over here on the second line, when I'm trying to add new score, I am once again adding floats and strings. Instead, I want that quiz score to be an integer so I can do some math with it.
To cast a value to an integer, we use the INT function. If I start with a string value like eight, nine, and I cast it to an integer, I am effectively just removing the quotation marks around it. So now I have 89. Do you see any problems here? Well, based on the previous quizzes, it seems that we support half points, like 89.5 might be a quiz score, and 89.5 is a float, not an integer.
If I enter 89.5 at the prompt, we get a value error because we're trying to cast the string 8 9.5 to an integer. This means that we probably shouldn't be casting to an integer. Instead, we should cast to a float. To cast a value to a float, we use the float function.
We might also move that typecast up here, so the value is already converted to a float before it's stored in the variable new score. Then we don't have to remember later to cast the value of new score to a float anytime we access it. Put that all together and we can cast values using the string function, the int function, and the float function.
Also, note, there is a bool function that converts values to booleans. If we try and cast values that just don't make sense, as you saw, we get a value error. This is like if I tried to cast the string banana to an integer or the string 9.9.9 to a float. It's just not possible in any meaningful way.
Like what should banana be as an integer? I don't know. Now what about casting a float to an integer? It turns out this just chops the decimal portion off. So 3.8 becomes three. We're taking the floor, or the lower closest integer. So next time you see a type error in your program, check the data types in your expressions and make sure they match strings with strings and integer and floats with integers and floats.
If you need those values to be a different data type, you can always cast them.
Video © Khan Academy, used under CC BY-NC-SA. Watch on Khan Academy. Puzzicle is not affiliated with Khan Academy.